We propose to combine pioneering developments in bioinformatics and comparative genomics with molecular biology expertise in a new model organism, the frog Xenopus tropicalis, to determine the genomic encryption of tissue-specific gene regulation, with emphasis on skeletal muscle and liver gene regulatory elements. In particular, we aim to elucidate sequence signatures, genomic location and the functional activity of transcriptional regulatory elements that drive tissue specific gene expression during embryonic vertebrate development. We propose to exploit unique computational and biological resources available at Livermore National Laboratory to develop a new generation of computational methods and tools capable of identifying and interpreting gene regulatory elements in anonymous noncoding genomic sequences. First, we propose to combine genome-wide gene expression profiling;vertebrate genome comparisons, and transcription factor binding site analysis to develop novel statistical methods for de novo prediction of skeletal muscle and liver tissue-specific regulatory elements in the human genome. Second, we will establish a high-throughput pipeline of in vivo experimental testing of the ehancer activity of predicted elements in frog, which is the only non-fish, sequenced vertebrate organism amnable to rapid and large-scale experimentation. We propose to test 550 predicted elements through the cource of 4 years. Through several rounds of computational prediction-experimentation validation, we aim to refine our methodology and to derive insights into the basic architecture of gene regulation in humans. We will also test the predicted tissue-specificity of 20 key skeletal muscle and liver transcription factor genes and 400 genes flanking predicted regulatory elements using in situ hybridization on frog embryos. Finally, we will create a publicly available resource for sharing the generated computational methods and tools along with experimental data with the broad research community.